Cognate Effect and Lexical Processing
in English-Spanish and Spanish-English
Bilinguals
Meredith Jocelyn Jane MCGREGOR*
Abstract
Cognates have served as a useful tool for investigating the bilingual lexicon in
many studies, but very little research has been carried out on different types of
cognates, specifically, partial cognates and their role in cross-linguistic effect.
The present study examines cognate effect in the speech production and acceptability judgment of two groups of highly proficient, late-onset English-Spanish
(n = 12) and Spanish-English (n = 12) bilinguals within a single-language (English)
context. The findings of two tasks, a production task, whereby participants were
asked to spontaneously produce synonyms to prompt words, and an acceptability judgment task of a variety of sentences including use of false and partial EnglishSpanish cognates are reported here, framed within non-selective, integrated models
of lexical representation. The results suggest a significant cognate effect in both bilingual groups in both tasks compared to their monolingual counterparts with, surprisingly, greater significance demonstrated from L2 to L1 influence, particularly in
production. These findings add to the growing support for semantic modulation at
the conceptual level of lexical processing in highly proficient bilinguals.
Keywords: bilingualism; cognate effect; lexical processing; Spanish; English.
*
University of Western Ontario, Ontario, Canada.
Corresponding author: [email protected]
Received: 2015-11-25 / Sent for peer review: 2016-01-27 / Accepted by peers: 2016-04-12 / Approved: 2016-06-24
To reference this article in APA style / Para citar este artículo en APA / Para citar este artigo
Mcgregor, M. (2016). Cognate effect and lexical processing in English-Spanish and Spanish-English bilinguals.
LACLIL, 9(1), 174-209. doi:10.5294/laclil.2016.9.1.8
174
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
El efecto de los cognados y el procesamiento léxico en
hablantes bilingües de inglés-español y de español-inglés
Resumen
El presente estudio examina el efecto de los cognados en la producción de un discurso y el juicio de aceptación de un grupo de 24 hablantes bilingües de inglés
y español y de español e inglés con un alto nivel de manejo de ambas lenguas
en un contexto donde se habla inglés únicamente. Los hallazgos de estas dos tareas se reportan aquí, y estas se enmarcan en modelos de representación léxica
no selectiva. Los resultados sugieren un efecto significativo de los cognados en el
grupo de participantes en las dos tareas en comparación con sus contrapartes
monolingües. Sorpresivamente, se aprecia una gran significancia relacionada con
la influencia de la segunda lengua en la primera lengua. Estos hallazgos apoyan la
modulación semántica que a nivel conceptual del procesamiento léxico sucede en
hablantes bilingües con altos niveles de manejo de las lenguas.
Palabras clave: bilingüismo; efecto de los cognados; procesamiento léxico; español.
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
175
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
O efeito dos cognatos e o processamento lexical em
falantes bilíngues de inglês-espanhol e de espanhol-inglês
Resumo
Este estudo analisa o efeito dos cognatos na produção de um discurso e o índice de aceitação de um grupo de 24 falantes bilíngues de inglês-espanhol e de espanhol-inglês com um alto nível de domínio de ambas as línguas num contexto
em que se fala inglês unicamente. Os achados de duas tarefas são relatados aqui,
as quais se delimitam em modelos de representação lexical não seletiva. Os resultados sugerem um efeito significativo dos cognatos no grupo de participantes
nas duas tarefas em comparação com suas contrapartes monolíngues. Observase uma grande significância relacionada com a influência da segunda língua na
primeira língua. Essas constatações apoiam a modulação semântica que, no âmbito conceitual do processamento lexical, se registra em falantes bilíngues com
alto domínio das línguas.
Palavras-chave: bilinguismo; efeito dos cognatos; processamento lexical; espanhol.
176
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
Introduction
Bilingual and multilingual individuals possess more than one set of lexical
representations from which to draw upon when faced with the demands
of speech production. As a result, there is always a cognate advantage in
speech production and recognition in bilinguals (Sherkina, 2003). However, as “cognate status is a continuum” (Szubko-Sitarek, 2015, p. 136), cognitive processing in bilinguals/multilinguals poses interesting questions,
especially regarding the conceptual level of lexical access (see Caramazza,
1997). Cognates are semantically, morphologically, phonologically (Amengual, 2012), pragmatically, and orthographically overlapping lexical items
drawn from categorically different lexicons. They are words that tend to
look, mean and sound the same in two or more languages, and they may
even be spelled the same way. Examples of orthographically identical cognates in English and Spanish are animal, canal, and escape. Although they
are pronounced differently (canal comes close: /kəˈnæl/ [English], and /
kaˈnal/ [Spanish], but is still not phonetically identical), on paper these
words are the same, and, semantically, they are more or less equal as well.
In its infinitive form in Spanish, escape diverges slightly as the reflexive
verb escaparse, but it carries so many similarities to the English word that,
even without a lot of context, one might be able to guess its meaning, as
with abundancia, famoso, and religión, known as semi-cognates (de Bot,
Cox, Ralston, Schaufeli, & Weltens, 1995, p. 4). And there are hundreds more
examples like these.1 However, something interesting about cognates is
that there is conceptual variation among them. False cognates (also known
as false friends) are words that may look and sound the same but that do
not function interchangeably as is the case in the word bigote – meaning
moustache, not bigot, as one might suspect, or asistir – meaning attend,
not assist. While identical and semi-cognates can be an asset to language
learners, these false cognates can have the opposite effect.
1 Johnston’s 1941 study, “Spanish-English Cognates of High Frequency”, looked at the
3022 most frequent words in Spanish and the 3000 most frequent words in English
and found one-third of them to be “related through Latin” and 50 others through other
sources, mostly Germanic roots (p. 405).
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
177
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
On the continuum between true cognates and false cognates exist
partial cognates (Child, 2010). Partial cognates are lexical items that sometimes work like cognates and sometimes, deceptively, do not. An example of
this is the Spanish noun convicción. It does mean conviction, as in a deeply
held belief, but its definition does not carry over into the additional meaning a guilty verdict as it does in English. Likewise, the Spanish word banco means bank in English, as in the financial institution, but it also means
bench. A Spanish speaker learning English would have to be careful with
this “false friend”, as referring to a park bench as a bank would cause confusion. A sub-category of partial cognates consists of those that, in spite of
possessing non-cognate definitions that do not translate directly, might be
used interchangeably among bilinguals without significant communication breakdown. Recall the reflexive verb cognate escaparse noted above. In
English, to escape denotes a sense of fleeing from some form of captivity.
In Spanish, the same is true, but this verb is also used to describe the act of
skipping school or cutting out early from work. There is considerable semantic overlap, but the degree to which an individual might accept: She escaped
early from work yesterday afternoon depends on one’s linguistic repertoire.
It might sound somewhat strange to a native English speaking monolingual, and the desired meaning may or may not be successfully conveyed.
To another native English speaker who is proficient in Spanish, it might not
seem so objectionable; it might not even be registered as unusual.
These partial cognates can cause what is known as cognate effect, which is when a non-target language plays a role in influencing a
target language. Due to their relatedness across languages, cognates can
affect language production and/or comprehension in bilinguals/multilinguals, but there are often inconsistencies. Individuals of varying language
profiles may differ in their use and assessment of such words, and this can
generate debate over whether there is redundancy in the way the bilingual
mind maps like lexical items.
Research questions and hypotheses
This study addresses the effect cognates could potentially have on lexical
access in highly proficient English-Spanish and Spanish-English bilinguals
178
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
through the analysis of data gathered from two distinct language tasks
including the production and acceptability judgment of lexical representations. The purpose is to examine the possibility of cross-linguistic influence in both L1 to L2 and L2 to L1 of these two languages, and the guiding
research questions of this study are as follows:
1.
What is the nature of cognate effect in highly proficient English-Spanish/Spanish-English bilinguals in spontaneous English speech production from an L1 to L2 versus an L2 to L1 approach?
2.
What is the nature of the way highly proficient English-Spanish/
Spanish-English bilinguals assess cognate lexical items in English
compared to their monolingual counterparts?
3.
What is the nature of the way English-Spanish/Spanish-English bilinguals both produce and assess true, partial, and false cognates?
In addressing the first research question, it was predicted that bilinguals would demonstrate involvement (i.e. cognate effect) between both
languages during the eliciting speech production task. This would be evidenced through proportionately higher employment of cognate words of English and Spanish when word frequency would dictate the likelihood of an
alternative word choice in the target language. Further, it was expected that
greater cognate effect would be found in the Spanish-English group working
in their non-dominant language (English), as per the Revised Hierarchical
Model (RHM) (Kroll & Stewart, 1994) which asserts that “word-to-concept”
mapping is stronger in the dominant language.
For the second research question, greater acceptance of problematic
lexical representations from the bilinguals compared to their monolingual
counterparts was expected, suggesting that language experience and use
informs conceptualization of lexical items. The hypothesis was that the
Spanish-dominant group would display the highest level of acceptability,
again as per the RHM, followed by the English-dominant group, and then
the monolinguals.
As for the third research question, it was hypothesized that there would
be no significant difference between the way the bilingual groups would
judge partial cognates but that they would accept them at a higher rate than
the monolinguals. However, it was expected that the Spanish-dominant
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
179
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
group, working in their L2, would judge false cognates to be more acceptable than either the English-dominant group or monolinguals, but that
the English-dominant group would still accept false cognates more liberally than the monolinguals, given their language experience.
Theoretical Background
Cognate effect
Cognates have been widely studied across a number of linguistic disciplines, as they represent overlap between two or more languages, as
demonstrated in the introductory examples. Lemhöfer, Dijkstra, and Michel
(2004) explain why cognates are so frequently and instrumentally used
in exploring the mental lexicon, stating, “Because they represent the lexical overlap between languages, they offer a straightforward way to tackle the question whether even in a language-exclusive setting, bilinguals
are influenced by their other language” (p. 587). This influence, or cognate
effect, has been shown to manifest itself in a number of different ways
and has been examined in several different linguistic sub-fields, including
psycholinguistic research in association and decision tasks in bilinguals
(Yudes, Macizo & Bajo, 2010), word recognition in bilinguals (Dijkstra, Van
Hell, & Brenders, 2014) and trilinguals (Lemhöfer, et al., 2004), word recognition and acquisition (de Groot, Borgwaldt, Bos, & van den Eijnden, 2002;
Midgley, Holcomb, & Grainger, 2011; Miller & Farr, 1939). Greater speed and
accuracy have been shown in naming cognates rather than non-cognate
items (Lemhöfer, et al., 2004) and even when task requirements included
only participants’ dominant language (Van Hell & Dijkstra, 2002). Further
evidence for cognate effect has been advanced in studies on voice onset
times in Spanish-English bilinguals of varying proficiencies (Amengual,
2012), citing more English-like pronunciation in the Spanish production of
cognates, i.e. phonetic interference with higher proficiency.
Cognates and non-cognates have also been an integral part of examining the processes of lexical decision, which is, generally, how quickly one
can discern cognates and non-cognates. Priming effects of cognate translations have been shown to demonstrate repetition in cognates as opposed
180
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
to non-cognates (Sánchez-Casas, Davis & Garcia-Albea, 1992), and interference from lexical neighbours and “meaning-related pairs” in translation
equivalent recognition testing has also been shown (Sunderman & Kroll, 2006). Other studies on lexical decision in monolinguals and bilinguals
involving cognate effect include: Costa, 2005; Costa, Miozzo & Caramazza, 1999; Costa & Santesteban, 2004, among others. These studies demonstrate how two (or more) mental lexicons may be stimulated at one time
within different processes of language processing. The process of inhibitory control (a process which effectively blocks a non-target language, theorized to occur at less proficient levels of language use – see Schwieter &
Sunderman, 2008; Schwieter & Sunderman, 2011; Schwieter, 2013), and language-specific mechanisms models (Schwieter & Sunderman, 2008) have
been explored, but research is nowhere near conclusive in this area (Finkbeiner, Almeida, Janssen & Caramazza, 2006). While results are not completely consistent, cognates have served as a useful tool for investigating
cross-linguistic influence, and considerable headway has been made in
this area of linguistics. In spite of on-going debate on a number of issues,
consensus seems to be drawing closer to cascaded, non-selective models
that support concurrent language activation in individuals proficient in
more than one language. A discussion on the progression of these models in support of this and how they tie into cognate effect in lexical access will follow.
Models of language processing and lexical access
“The biggest challenge for a model of the bilingual lexicon in general is
that it should account both for cross-language effects and the ability of
bilinguals to use one language at a time” (Sherkina, 2003, p. 139). Indeed,
explaining bilingual language processing is a complicated undertaking.
However, several different models of language processing have been put
forth, and many of them account for this persistent cognate effect. What
is more important to the present study is that some of the modelling can
account for multiple manifestations of single lexical items, suggesting
that conceptual overlapping, like what we see in cognates (including partial cognates), can be reconciled. These models include those in support
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
181
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
of integrated lexical storage and non-selective access, as well as those that
can account for individual, experiential variables of language users.
In terms of lexical access, it is generally agreed that there are two
stages of processing: the conceptualization of the semantic/syntactic lexical representation, i.e. lemma, and then the selection of its corresponding
phonetic representation, i.e. lexeme, as per the Dual Stage Access Model
(Caramazza, 1997). Though it is described in binary terms in this model, it
is a truly nuanced process. Both “selective” (meaning not simultaneously
activated) and “non-selective” (meaning simultaneously activated) models
have been put forth, but there is still controversy over whether one lexicon
may ever be selected for discretely (see Costa, Heij, & Navarrete, 2006), or
whether more than one language is always activated as per the Non-Selective Access Hypothesis (Lemhöfer, et al., 2004; Morsella & Miozzo, 2002).
The latter, non-selective access, has come to be more accepted as a result
of more sophisticated research in cognitive processing (Van Hell and Dijkstra, 2002), and as a result of a number of other studies supporting this theory (Cuppini, Magosso, & Ursino, 2013; de Groot, Delmaar, & Lupker, 2000;
Dunlap, Chen, Zhou, & Yang, 2010, among others). There have been studies suggesting selective access (i.e. no influence from the non-target language), but it has been speculated that proficiency may have been too low
in the participants (Van Hell & Dijkstra, 2002), and Dijkstra et al. (2014) warn
that null cognate effect findings in behavioural data does not necessarily
mean that there is no cognitive effect. Kroll and de Groot (2005) compare
language selection to register selection arguing that just as monolinguals
do not select for formal language versus slang or “taboo words” there is no
need for bilinguals to select for words of a particular language (p. 305). In
this view, all language users are basically the same. Having access to another mental dictionary (mental lexicon) through knowledge of a second
language is no different from having access to different variations of one
single language. This notion is supported by many other scholars (Antoniou, Best, Tyler & Kroos, 2010; de Bot & Jaensch, 2015), who call for a more
integrated approach to language processing, treating all languages as collective entities rather than separate ones in the brain.
In addition to a non-selective view, addressing how users of two or
more lexicons with overlapping lexical representations (i.e. cognates) can
182
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
function in one independent, target language has led to debate over either
separate or integrated lexical storage of lexical representations. According to Kroll and de Groot (2005), “It is generally assumed that, during conceptual processing, not only the semantic representation of the intended
concept but also those representations of semantically related concepts
are activated to some degree” (p. 309). This remark is important to the current study because it means that in language production, conceptual processing also involves discerning between related concepts (i.e. cognates)
before moving on to the next step: lexical selection. Basically, the lexicon
of one language is not completely separate from another within this conceptualization stage. If this is true, then it would make sense for a highly
proficient bilingual to produce and/or accept lexical representations with
unclear conceptual boundaries (like those in partial cognates) in a way that
monolinguals simply could not. This idea of semantically related concepts
and how they are stored/accessed is supported by a number of models. The
One-Store Hypothesis advocates for a single, integrated memory system
of lexical representation, the Two-Store Hypothesis posits a separate system, and the Three-Store Hypothesis suggests that two or more languages
are “differentially connected” in one “conceptual-experiential information store” (Szubko-Sitarek, 2015, pp. 67–68). Another well-known model
in support of the integrated viewpoint is the Bilingual Interactive Activation + (BIA+) Model (Dijkstra &Van Heuven, 2002), successor to the original BIA Model (Dijkstra &Van Heuven, 1998). While this model is not able
to account for language development (Basnight-Brown, 2014), it features
an integrated word identification system with language nodes utilized
to specify which language a lexical item belongs to and that word recognition is only somewhat influenced by language membership. However,
different types of studies have produced results supporting each of these
hypotheses. In particular, the integrationist view of lexical representation
has been supported by studies citing cross-linguistic influence (Szubko-Sitarek, 2015). However, Sherkina (2003) points out that in the cognitive processing of cognates, conceptual representation and sound inventories do
not necessarily come from single lexical nodes, but some processing models (i.e. the BIA+ Model) assume this, proving problematic.
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
183
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
In addition to the aforementioned, static models of lexical storage and
access, there is increasing support for the view that language processing
and lexical representation may be more fluid, dependent on individual variables (Dijkstra & van Heuven, 2002). This means we should consider how
acts of “knowing, learning and using an additional language can interact
with lexical access (Altarriba & Heredia, 2014, p.9). For example, Cuppini,
et al. (2013) suggest that higher language proficiency results in greater similarity in semantic systems, meaning that interlingual relationship systems
may change over time, and, consequently, that individual language users’
processing may be more dynamic than what the aforementioned models
account for. This is supported in the previously mentioned RHM and L1 to
L2 connections and echoed by Szubko-Sitarek (2015) stating that “with the
increasing proficiency, stronger and stronger semantic links between concepts and L2 items are developed” (p.79). Something important to highlight
here is that these are non-selective, integrated, proficiency-based theories
but that the RHM still presents a scenario whereby there are two separate
lexicons. As a result, Desmet and Duyck (2007) proposed another revision
of the RHM, one that depicted one lexicon with an independent semantic
store. These models do account for the way in which lexical items may be
conceived of and produced, but there are always individual differences in
how words are conceptualized, selected for. To explain this, while also taking individual variability into account, we turn to usage-based, exemplar
theory models. According to Bybee (2013) citing her Usage-Based Theory
from 1985, “representations are strengthened by repetition, making them
easier to access” (p. 59). Bybee (2001) plainly states, “Experience affects representation” (p. 6). These concepts of repetition and experience are fundamental to how individuals access lexical resources and may be applied to
L1, L2, or Ln users. How this applies to an understanding of lexical representation in bilinguals may be further understood by Exemplar Theory.
According to Gahl and Yu (2006), redundancy in mental representations
is central to exemplar models in linguistic theory, meaning lexical representations are neither conceptualized nor selected for in isolation. Exploring Exemplar Theory and the lexical/phonetic interface, Amengual (2012)
suggests that “all the possible phonetic manifestations and lexical mean184
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
ings are stored with each unit (word), in addition to different contextual
variations and realizations” and that each lexical item “is stored with re-
dundant phonetic, semantic, and contextual information represented in
memory by a “cloud” of remembered exemplars which directly reflect a
speaker’s experience with the specific word in both production and per-
ception” (p. 527). This means that semantically overlapping lexical items
are concurrently activated, stored together, not in isolation, and that one
particular user may store different items differently as per experience and
usage. This and the other models in support of integrated lexical storage
and non-selective access, as well as those that can account for individual,
experiential variables of language users can account for multiple manifestations of single lexical items, suggesting that conceptual overlapping, like
what we see in cognates (including partial cognates) and cross-linguistic
influence can be better understood.
Method
This study is comprised of two tasks, both carried out with participants
placed in a single-language context (Green & Abutalebi, 2013) whereby
only the use of English was requested. In the first task, participants were
asked to spontaneously produce English synonyms to a list of 45 English
words. The task was carried out verbally, and participants were not shown
the transcribed list of words, so that there would be no visual/orthographic priming effect (Van Hell & De Groot, 2008). Words in Task 1 were pre-
sented in isolation, without the benefit of being embedded within a larger
matrix sentence for semantic priming, and the participants were asked
to respond one at a time. Responses were audio recorded in order to elicit
natural, immediate responses, in lieu of formal time constraint, as to avoid
“strategizing” (Szubko-Sitarek, 2015, p. 166).
Target words were selected based on known synonyms with cog-
nateness between English and Spanish, and filler control words were also
included. Word forms included verbs, adjectives and nouns. For example,
some of the words included were: to reply, friendly, and environment. A
cognate synonym for these examples would include: to respond, amicable,
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
185
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
and ambiance. Fillers with no intended cognate relationship between English and Spanish such as to wash, goofy, and dream were also included to
serve as control words. Whether the participants would more often produce these cognate synonyms in spite of their being less frequently used
English words, or whether they would produce entirely dissimilar items
was of interest here. The idea was that uttering less frequent cognate synonyms would demonstrate possible effect from the non-target language
(Spanish). The completed list can be found in Appendix A.
In the second task, participants were asked to rate their level of acceptability on a scale of 1–5 of 60 sentences, listed in full in Appendix B,
containing four different types of variables: cognates, preposed adjectives,
preposed adverbs and inverted noun clauses. As the focus of this study is
cognate effect, only the cognate/semantically overlapping tokens will be
analyzed and discussed here. These include partial and false cognates, as
described in the introduction of this paper. The scale markers were defined as follows:
1.
Unacceptable; meaning is completely unclear; I would never express
myself in this way.
2.
Somewhat unacceptable; meaning is somewhat unclear; I am unlikely to express myself in this way.
3.
Satisfactory; I probably wouldn’t say it this way, but I understand
the meaning just fine.
4.
Somewhat acceptable; meaning is somewhat clear; I might express
myself in this way.
5.
Acceptable; meaning is completely clear; I would most likely express
myself in this way.
For both types of cognates (partial and false), there was always a
paired equivalent sentence featuring an alternative, standard word in place
of the partial/false cognate. It is important to note that the sentence pairs
in Task 2 were not presented together. All 60 sentences were randomized,
so participants were not asked to compare pairs beside one another, but
rather assess them one at a time, in no particular order. This first pair exemplifies the partial cognate presented in 1a., and its more standard or correct synonym introduced in 1b.
186
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
(1)
a. The professor presented herself to the class on the first day.
b. The professor introduced herself to the class on the first day.
While the meaning of present in this context is very similar to that
of introduce, 1b. is the more standard choice. This next pair exemplifies the
false cognate gracious in 2a and the more appropriate word choice funny in 2b.
(2) a. My mom is always telling jokes; she is so gracious.
b. My mom is always telling jokes; she is so funny.
Pairs like these were included among the total 60 sentences, and
like Task 1, participants were asked to respond without hesitation, rating
each sentence based on their first impression. Unlike Task 1, participants
were able to see and read the target sentences, so they did have the benefit of both sentence context and visual (orthographic) cues to make their
acceptability judgments.
This study features 36 participants in three participant groups with
a ratio of 23 females to 13 males. A number of adult, highly proficient English-Spanish (n=12) bilinguals, highly proficient Spanish-English (n=12)
bilinguals, and monolingual Anglophones (n=12) were invited to carry
out two tasks with all interaction/instructions spoken in English. Prior to
testing, participants were asked to fill out a detailed language proficiency
background questionnaire citing, among other details: the age they began
to learn their L2, hours of L2 use per week, self-reported proficiency and the
level of importance their L2 holds for them. A number of the English-Spanish bilinguals cited some formal schooling in French but expressed superior
proficiency in and much more frequent use of Spanish, which was important to this study. Table 1 summarizes the key information gathered through
this questionnaire (Table 1).
As seen in Table 1, the first group consisted of all late English-Spanish bilinguals raised in English-speaking Canadian/American homes with
an average age of Spanish onset of 16.7. The third group, the control group,
consisted of all monolingual, Anglophone individuals ranging in age from
young adult to early senior, who, in spite of some reported French language schooling as part of mandatory Ontario French curriculum as children, cited zero contact with French language and minimal to no ability
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
187
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
Table 1. Bilingual background information
Participant
Group
Average
Self-Reported
Importance
Average age
Weekly Use
L2 language
rating
of L2 Onset
of L2 (in
proficiency
1=None;
M(SD)
hours)
1=None; 6=Native
10=Utmost
M(SD)
M(SD)
M(SD)
English-Spanish
16.7 (8.1)
12.1 (12.2)
4.3 (0.5)
9.3 (1.1)
Spanish-English
15.0 (7.5)
30.3 (15.8)
4.1 (0.7)
9.7 (0.7)
Note. All members of this group reported active, weekly use of Spanish (in spite of living in
English-speaking countries), high or near-native proficiency in Spanish and rated Spanish as
being of great importance in their lives. The second target group consisted of all late SpanishEnglish bilinguals with an average age of English onset of 15.0 who were formally trained in
English as a second language and who were raised in Spanish-speaking countries in Spanishspeaking homes but who are all currently living in English-speaking Canada, with the exception
of one who now resides in the United States. All members of this group reported consistent and
regular use of both languages on a weekly basis.
to communicate in any language other than English. All participants had
attained some level of post-secondary schooling, and, with the exception
of four, all cited possession of at least some level of university education;
those four had completed alternative post-secondary schooling. Those who
cited university studies had successfully graduated or are currently in the
process of doing so, and a number of them had already completed or are
currently pursuing post-graduate studies.
Results
The descriptive analysis for Task 1 revealed interesting findings between both
the bilinguals as a whole and the monolingual control group, and between
the two bilingual groups themselves. Figure 1 reveals the overall averages of
cognates produced by each group, for each part of speech (verbs, adjectives
and nouns), showing that there was a comparable rate of cognate effect in
both bilingual groups.
Participants from the English-dominant group produced the highest
average of cognate words across all three categories and across all three
188
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
Figure 1. Task 1: Overall cognate averages produced
for each part of speech, per group in percentages
groups with an overall mean of 49.7% (sd = 0.31). Within the verbs category specifically, the English-dominant group produced a mean of 54.71%
(sd = 0.3) and almost an equal number of cognates within the noun category with a mean of 54.24% (sd = 0.37), and within the adjectives category,
they produced an average cognate mean of 40.15% (sd = 0.25). While the Spanish-dominant group also produced a considerable number of cognate items,
the overall average was significantly lower than that of the English-dominant group with an overall mean of 42.52% (sd = 0.29). Within this group,
the category that revealed the most cognate words was the nouns category
(mean = 50.63%, sd = 0.31), which is comparable to the mean produced by the
English-dominant group. Within the verbs and adjectives categories, however, the average number of cognates was 39.39% (sd = 0.31) and 37.55% (sd
= 0.25), respectively. Calculating the average number of cognates produced
from both bilingual groups, the overall resulting mean is 46.11% (sd = 0.3) in
contrast to the overall mean of 23.62% (sd = 0.2) produced by the monolingual group. Like the Spanish-dominant group, the monolinguals produced
the highest number of cognates within the nouns category (m = 28.79%,
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
189
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
sd = 0.23). The second highest percentage of cognate words came from the
adjective category (m = 20.52%, sd = 0.21) for the monolinguals, and within the verbs category the mean was 21.57% (sd = 0.17).
Figure 2, Figure 3, and Figure 4 provide breakdowns of overall cognate
averages and cognate responses per prompt word for each group: verbs,
adjectives, and nouns, respectively.
In many cases, higher cognate rates are seen in bilinguals for the
same prompt, demonstrating a similar nature of cognate effect. In order
to examine this effect more carefully, however, it is necessary to reference
word frequencies within the English language used in Canada in order to
see whether or not the lexical items produced by the participants represent words that are simply just more frequently used, or whether they are
less frequent words, and in the case of the bilinguals, selected for as a result of cognate effect coming from the non-target language. A complete
list of the most frequently produced lexical items from Task 1 as compared
to English word frequencies can be seen in Table 2. The “Prompts” column
on the left hand side lists the words for which participants were asked to
Figure 2. Verb cognates produced for each prompt,
per group, in percentages
190
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
Figure 3. Task 1: Adjective cognates produced for each prompt,
per word, per group, in percentages (%)
Figure 4. Task 1: Noun cognates produced for each prompt,
per word, per group, in percentages (%)
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
191
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
spontaneously provide synonyms, and the most frequent responses by each
of the three test groups are provided. The “Freq” columns indicate how frequently each word appears in the Corpus of Canadian English (Strathy) (n.d.)
of 50 million words. The higher the number, the more frequent the word.
Freq
Most frequent responses
by Spanish-dominant
Group
Freq
Most frequent responses
by Monolingual Group
226
Respond
2722
Answer
2997
Respond
2722
485
Arrive
1266
Appear
5149
Arrive
1266
Introduce
1608
Present
1704
Present
1704
Greet
181
Signal
328
Indicate
3283
Show
10053
Sign
1454
Research
3032
Investigate
1310
Investigate
1310
Study
2744
Say
31875
Speak
5660
Speak
5660
Speak
5660
Bother
537
Annoy
39
Annoy
39
Annoy
39
Apply for
426
Solicit
138
Submit
1021
Submit
1021
Chat
212
Talk
6517
Talk
6517
Talk
6517
Remind
694
Remember
5793
Remember
5793
Repeat
1295
Different
22361
Distinct
2112
Distinct
2112
Opposite
876
Freq
Most frequent responses
by English-dominant
Group
Reply
Show up
Prompts:
Freq
Table 2. Most frequent word responses in Task 1 and Canadian
English word frequencies
Friendly
1137
Amicable
34
Nice
2556
Kind
624
Weak
1723
Not strong
60
Soft/Tired
1726
Feeble
118
Current
10282
Actual
3371
Up-to-date
228
New/Now
55900
Vague
547
Unclear
1327
Not clear
562
N/A
Unlikely
1586
Improbable
117
Different
22361
Uncommon
408
Healthy
3234
Strong
8657
Good shape
99
Fit
1451
Sensitive
1886
Sensible
443
Emotional/
Sensible
1456
Emotional
2468
Funny
1015
Comical
46
N/A
Hilarious
110
Weird
324
Strange
2013
Strange
2013
Strange
2013
Behaviour
4305
Comportment
6
Actions
3148
Actions
3148
Tradition
2923
Custom
699
Culture
8294
Custom
699
192
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
Freq
53
Most frequent responses
by Monolingual Group
6192
Folder
Freq
Concern
Most frequent responses
by Spanish-dominant
Group
6299
Freq
Freq
Goal
Most frequent responses
by English-dominant
Group
Prompts:
MCGREGOR
Objective
1651
Objective
1651
Aim/Plan
6489
Worry
235
Preoccupation
268
Worry
235
File
1827
File
1827
Binder
259
Relative
331
Family
21495
Family
21495
Family
21495
Attempt
2925
Try
248
Try
248
Try
248
Greeting
209
Salutation
10
Salutation
10
Hello
429
Review
8220
Revision
542
Revise
196
Check
689
Environment
10044
Ambiance
35
Context
5642
Surroundings
367
MEAN
4167.3
2445.3
4005
4170
SD
6994
4230
5621
10766.3
As seen in the above table, the verb respond was given by the English-dominant group as a synonym to the verb reply most often (m = 66.67%).
The synonym answer was provided in all other cases (m = 33.33%). Citing the
Corpus of Canadian English (Strathy) (n.d.), however, the verb answer appeared 2997 times in the 50-million-word corpus, or slightly more times
than the verb respond appeared, which was 2722 times. When factoring
in answer as a noun, however, its frequency almost triples, making it a
highly common word within Canadian English. This suggests that the English-dominant bilinguals were choosing the less frequent English-Spanish cognate respond more often than the more frequent word in English
answer. The Spanish-dominant groups had a near opposite effect, producing answer much more frequently as a synonym to reply (m = 75.0%) with
respond as a response only 17% of the time. Interestingly, the monolinguals
also produced respond (m = 58.3%) more frequently than answer (m = 25%),
so in this case it seems inconclusive whether or not the English-dominant
bilinguals were influenced by their L2 Spanish lexicon or whether respond
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
193
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
is just a popular synonym for reply. However, there is a marked difference with the responses to the verb research. Study appears in the Corpus
of Canadian English (Strathy) (n.d.) 2744 times as a verb, while investigate
appears only 1310 times. Interestingly, both bilingual groups more often
provided investigate, which has a Spanish cognate, as a synonym, while
not one monolingual produced that verb.
Another finding worth mentioning appeared in the nouns category.
For behaviour, the word comportment was produced by the English-dominant group as a synonym 41.67% of the time, which was more often than
any other response. Only 17% of the Spanish-dominant bilinguals produced
this response, with quite a variety among the other responses. The monolingual group did not produce this word. The word comportment only appears in the Corpus of Canadian English (Strathy) (n.d.) 6 times out of 50
million words, suggesting that the English-dominant group may be influenced by their use of Spanish and the cognate comportamiento. Similarly,
ambiance was produced by the English-dominant participants 33.33% of
the time for the prompt environment. Spanish-dominant participants produced ambiance only one time (8%), while this word was never produced
by the monolinguals. Like comportment, ambiance appears rarely in the
Corpus of Canadian English (Strathy) (n.d.), only 35 times. Thus, in many cases words with lower English frequency appeared with greater frequency
among bilinguals, especially the English-dominant bilinguals, providing
evidence for possible cognate effect, and an even greater effect in participants working within their dominant language. Furthermore, by calculating the average overall frequency of words produced by each group, the
Spanish-dominant and Monolingual groups produced almost an identical rate of frequency, even though their specific word choices were different in many cases. The English-dominant participants produced a much
lower average word frequency at 2445.3 per 50 million words compared
to the other two groups producing close to double that in the 4000 words
per 50 million range.
The responses from the number of items not listed in the relatively
small size of each individual group. Analysis was carried out for the three
individual groups on four different sentence types including sentences
containing partial cognates, sentences containing false cognates, and sen194
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
tences representing their “correct” respective pairings. Table 3 displays the
between-group analyses for each sentence type, as indicated in the first
column, and whether or not a significant effect was detected, indicated in
the third column.
Table 3. Task 2: Between-groups analyses per sentence type
Sentence Type
Partial Cognates (PCs)
Corrected PCs
False Cognates (FCs)
Corrected FCs
Group
p value
Eng-dominant – monolinguals
(p = 0.235)
Sp-dominant – monolinguals
(p = 0.082)
Eng-dominant – Sp-dominant
(p = 0.212)
Bilinguals – monolinguals
(p = 0.382)
Eng-dominant – monolinguals
(p = 0.006)
Sp-dominant – monolinguals
(p = 0.204)
Eng-dominant – Sp-dominant
(p = 0.024)
Bilinguals – monolinguals
(p = 0.032)
Eng-dominant – monolinguals
(p = 0.22)
Sp-dominant – monolinguals
(p = 0.003)
Eng-dominant – Sp-dominant
(p = 0.079)
Bilinguals – monolinguals
(p = 0.024)
Eng-dominant – monolinguals
(p = 0.006)
Sp-dominant – monolinguals
(p = 0.438)
Eng-dominant – Sp-dominant
(p = 0.031)
Bilinguals – monolinguals
(p = 0.061)
As can be seen from Table 3, no significant effect was found between
any group within the partial cognate sentences, except between the Spanish-dominant and monolingual groups (p = 0.08, u = 57), which was reaching significance (df = 22). However, within the sentences containing correct
lexical items in place of the partial cognates, there was significant effect between the English-dominant and monolingual groups (p = 0.006, u = 35.5),
between the Spanish-dominant and monolingual group (p = 0.02, u = 46),
and between the bilingual groups combined and the monolinguals
(p = 0.032, u = 102.5) (df = 34). No significant effect was found between the
Spanish-dominant group and the monolinguals (p = 0.2, u = 68).
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
195
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
Analysis of the groups and their judgments of the sentences containing false cognates also produced significant effect in some cases. Specifically, there was an effect found between the Spanish-dominant group and
the monolinguals (p = 0.003, u = 30), between the bilinguals as a whole
and the monolinguals (p = 0.02, u = 98), and analysis between the two bilingual groups showed a reaching significance effect (p = 0.08, u = 56.5).
There was no effect found for sentences with false cognates between the
English-dominant and monolingual groups (p = 0.22, u = 69). The sentences containing correct lexical items in place of the false cognates also
produced significant effect between some groups, namely between the
Spanish-dominant and monolinguals groups (p = 0.006, u = 35) and between the two bilingual groups (p = 0.03, u = 48). There was a nearly significant effect between the bilinguals as a whole and the monolinguals
(p = 0.06, u = 112), but no effect between the Spanish-dominant group and
the monolinguals (p = 0.44, u = 81).
Looking at the overall judgment means, using the 1–5 scale, the English-dominant group averaged 3.9 (sd = 1.04), the Spanish-dominant group
averaged 3.9 (sd = 0.79) as well, and the monolinguals averaged 3.6 (sd = 1.05).
In all cases, each group judged the “correct” sentences to be more acceptable
than those containing partial or cognate words, but as the statistical analysis (above) shows, there was significant variation throughout. The fact
that there was greater statistical significance in the so-called “correct” sentences is of interest because, in theory, the native English speakers should
have accepted them in a more similar way.
Discussion
The results from Task 1 indicate that cognates were produced in greater
numbers for the English-dominant group than the other two groups. This
finding runs contrary to the original hypothesis, which predicted that
there would be greater cognate effect from Spanish demonstrated within the Spanish-dominant group working in their L2. The original hypothesis was based on the RHM (Kroll and Stewart, 1994), which suggests that
“word-to-concept” mapping is stronger in the dominant language, but the
results here do not support that. A significant number of partial cognates
196
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
(and even false cognates) were produced by the English-dominant bilingual group, suggesting that L2 interference seemed to be in effect. This is
an interesting finding given that the English-dominant group was working
within a native language interactional context. We do know that stronger
semantic links are developed as proficiency increases (Van Hell & Dijkstra,
2002), so this could account for these highly proficient English-dominant
bilinguals being influenced by their L2. However, this does not account
for the finding that Spanish cognate effect also appeared in the equally proficient Spanish-dominant English users. This latter group seemed
to be highly influenced by frequently used words in English, which also
suggests L2 influence, but their rates of Spanish cognate words (including
false cognates) was still much higher than the monolinguals. This variability between the bilinguals points towards usage and exemplar models
presented earlier which suggest that individual factors such as repetition
and experience can impact how lexical representations may be conceptualized and, subsequently, produced. It is possible that these Spanish-dominant English bilinguals, living in an English-speaking environment,
are more influenced in their L2 by the input they receive in their L2,
rather than by their L1, but this contradicts the RHM in that it suggests
the most dominant language should still prevail. It is important, however, to acknowledge the subjectivity within Task 1 and the selection
of the target words. While they were selected for their known synonyms with cognateness between English and Spanish, a better approach may have been to precede the current study by following one of
two previous methods: a similarity rating task (De Groot & Nas, 1991), or a
translation-elicitation task (Kroll & Stewart, 1994) to assess the cognates
beforehand. Unfortunately, such measures were not taken prior to carrying
out the present study and were, therefore, not controlled for. A future replica of this study is needed to better confirm the current results with more
careful attention to cognate designation. In spite of this, it is believed that
the differences in production across the groups were significant enough
to account for any potential margin of subjectivity.
The results from Task 2 also suggest that these independent groups
are different in the way they negotiate lexical items, as partial cognates,
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
197
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
false cognates and even perfectly standard, non-cognate lexical representations were judged differently. Interestingly, in comparing the results between the two tasks, the way cognates were produced and the way that
they were perceived were different. While the production task showed the
English-dominant group producing much higher rates of the different types
of cognates as compared to their monolingual counterparts, the judgment
task showed no significant difference in how these two groups perceived
either partial or false cognates. This suggests that there was no statistically significant cognate effect in the English-dominant bilinguals, counter
to prior research (Sherkina, 2003) claiming that there is always a cognate
effect. More surprising still, is the fact that there was a significant effect
in the way these two groups (both English natives) assessed the “correct”
sentences. The reverse was true for the Spanish-dominant participants as
compared to the monolinguals in the second task, as there was more significant effect found in the partial and false cognates than in the correct
sentences. This suggests that these two groups differed in the way they
judge cognates but not in the way they judged correct sentences, which
means there was a cognate effect but that their L2 competency was strong
enough to recognize correct sentences like a monolingual. This finding for
the semantically overlapping partial and false cognate items supports the
RHM and the idea that L1 conceptualization should dominate. This matches the Task 1 results in one way but not in another: the Spanish-dominant
group may have been influenced by their L1 in terms of cognates in some
measure, but not so in terms of how they assessed correct sentences. Such
results suggest that language experience and usage may have given them
an edge in production but not in interpretation.
The results from both tasks support previous work on cognate effect and cross-linguistic influence in part, but also support exemplar
and usage-based theories. While it is not entirely conclusive, this provides further evidence for the notion that knowing more than one language interacts with the way lexical items are processed at least in some
ways. The theoretical implications here are that languages may be activated concurrently in bilinguals, as per the non-selective models, and
that selection of lexical items may be integrated whereby like concepts
198
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
are associated in their use, but based on the findings in this study, it is
not possible to speculate on where or how they are stored in the brain.
However, the use of cognates in this study as a tool has allowed for some
further exploration into semantic modulation at the conceptual level of
lexical processing and has contributed to some further support for the
idea that second language knowledge can influence dominant L1 performance. Finally, due to the variability found in the results, notwithstanding any inherent flaws to the study itself, some measure of individual
language experience must be acknowledged in the way models of lexical processing are designed.
Conclusion
To conclude, this study has looked at cognate effect and lexical processing
in highly proficient English-Spanish and Spanish-English bilinguals. It has
been able to provide further evidence for cross-linguistic influence through
cognate effect and discussion of lexical items both in production and in the
way they are judged. While this study cannot account for all idiosyncrasies found within the analyses, it has demonstrated some clear differences in the performances of the three participant groups. English-dominant
bilinguals demonstrated influence from their L2 in the production task but
showed variation in the judgment task as compared to the other groups.
This shows that at high levels of language competency, an L2 can influence and even interfere with dominant, native language production. The
Spanish-dominant group seemed to demonstrate some influence from
both their L1 and their L2 in that they produced a significant number of
cognate words but also many highly frequently used English words; they
also showed both L1 and L2 influence in the judgment task with significant
differences between their judgments in partial and false cognates compared to the monolinguals. This suggests input and language context to be
a powerful influence. The bilinguals combined generally showed an overall significant difference compared to the monolinguals, supporting the
notion that knowledge of a foreign language does inform knowledge of a
native language and vice versa. In particular, given the semantic overlapping of the lexical items featured in this study, it may be suggested that
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
199
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
at the conceptual level of lexical processing, there is evidence for integrated storage of like lexical representations.
References
Altarriba, J., & Heredia, R. R. (2014). Introduction: Bilingual memory
representation. In R. R. Heredia & J. Altarriba (Eds.), Foundations
of bilingual memory (pp. 3–10). New York, NY: Springer New York.
doi:10.1007/978-1-4614-9218-4
Amengual, M. (2012). Interlingual influence in bilingual speech: Cognate
status effect in a continuum of bilingualism. Bilingualism:
Language and Cognition, 15(3), 1–14. doi:10.1017/S1366728911000460
Antoniou, M., Best, C. T., Tyler, M. D. & Kroos, C. (2010). Language context
elicits native-like stop voicing in early bilinguals’ productions in
both L1 and L2. Journal of Phonetics, 38, 640–653.
Basnight-Brown, D.M. (2014). Models of lexical access and bilingualism. In
R. R. Heredia & J. Altarriba (Eds.), Foundations of bilingual memory
(pp. 85–107). New York, NY: Springer New York. doi:10.1007/9781-4614-9218-4
Caramazza, A. (1997). How many levels of processing are there in lexical
access? Cognitive Neuropsychology, 14, 177–208.
Child, J. (2010). Introduction to Spanish translation (3rd ed.). Lanham, MD:
University Press of America.
Corpus of Canadian English (Strathy). (n.d.). [Data set]. Retrieved from http://
corpus.byu.edu/can/
Costa, A. (2005). Lexical access in bilingual production. In J. Kroll & A. de
Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches
(pp. 308–325). New York, NY: Oxford University Press.
Costa, A., Heij, W. L. & Navarrete, E. (2006). The dynamics of bilingual
lexical access. Bilingualism: Language and Cognition, 9(2), 137–151.
doi:10.1017/S1366728906002495
Costa, A., Miozzo, M. & Caramazza, A. (1999). Lexical selection in bilinguals:
Do words in bilingual’s lexicon compete for selection? Journal of
Memory and Language, 41, 365–397.
200
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
Costa, A. & Santesteban, M. (2004). Lexical access in bilingual speech
production: Evidence from language switching in highly proficient
bilinguals and L2 learners. Journal of Memory and Language, 50(4),
491–511. doi:10.1016/j.jml.2004.02.002
Cuppini, C., Magosso, E. & Ursino, M. (2013). Learning the lexical aspects of a
second language at different proficiencies: A neural computational
study. Bilingualism: Language and Cognition, 16(2), 1–22. doi:10.1017/
S1366728911000617
de Bot, K. Cox, A., Ralston, S., Schaufeli, A., & Weltens, B. (1995). Lexical
processing in bilinguals. Second Language Research, 11(1), 1–19.
doi:10.1177/026765839501100101
de Bot, K., & Jaensch, C. (2015). What is special about L3 processing?
Bilingualism: Language and Cognition, 18(2), 130–144.
Bybee, J. (2001). Phonology and language use. Cambridge, UK: Cambridge
University Press.
Bybee, J. (2013). Usage-based theory and exemplar representations of
constructions. In T. Hoffmann & G. Trousdale (Eds.), The Oxford
handbook of construction grammar, (pp. 49–69). New York, NY: Oxford
University Press. doi:10.1093/oxfordhb/9780195396683.013.0004
De Groot, A. M. B., Borgwaldt, S., Bos, M. & van den Eijnden, E. (2002). Lexical
decision and word naming in bilinguals: Language effects and task
effects. Journal of Memory and Language, 47(1), 91–124.
De Groot, A. M. B., Delmaar, P. & Lupker, S. J. (2000). The processing of
interlexical homographs in translation recognition and lexical
decision: Support for non-selective access to bilingual memory. The
Quarterly Journal of Experimental Psychology Section A: Human
Experimental Psychology, 53(2), 397–428. doi:10.1080/713755891
De Groot, A. M. B., & Nas, G. L. J. (1991). Lexical representation of cognates
and noncognates in compound bilinguals. Journal of Memory and
Language, 30, 90–123.
Desmet, T., & Duyck, W. (2007). Bilingual language processing. Language
and Linguistics Compass, 1(3), 168 – 194. doi:10.1111/j.1749818X.2007.00008.x
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
201
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
Dijkstra, T., Van Hell, J. G. & Brenders, P. (2014). Sentence context effects in
bilingual word recognition: Cognate status, sentence language,
and semantic constraint. Bilingualism: Language and Cognition,
18(4), 1–17. doi:10.1017/S1366728914000388
Dijkstra, A. & van Heuven, W. (1998). The BIA–model and bilingual word
recognition. In J. Grainger & A. Jacobs (Eds.), Localist connectionist
approaches to human cognition (pp. 189–225). Mahwah, NJ:
Erlbaum.
Dijkstra, A. & van Heuven, W. J. B. (2002). The architecture of the bilingual
word recognition system: From identification to decision.
Bilingualism: Language and Cognition, 5, 175–197.
Dunlap, S., Chen, B., Zhou, H. & Yang, M. (2010). Language nonselective access
to phonological representations: Evidence from Chinese-English
bilinguals. The Quarterly Journal of Experimental Psychology,
63(10), 2051–2066. doi:10.1080/17470211003718705
Finkbeiner, M., Almeida, J., Janssen, N. & Caramazza, A. (2006). Lexical
selection in bilingual speech production does not involve language
suppression. Journal of Experimental Psychology: Learning, Memory,
and Cognition, 32(5), 1075–1089. doi:10.1037/0278-7393.32.5.1075
Gahl, S., & Yu, A. C. L. (2006). Introduction to the special issue on exemplarbased models in linguistics. The Linguistic Review, 23, 213–216.
Green, D., & Abutalebi, J. (2013). Language controls in bilinguals: The
adaptive control hypothesis. Journal of Cognitive Psychology, 25(5),
515–530. doi:10.1080/20445911.2013.796377
Johnston, M. C. (1941). Spanish-English cognates of high frequency. The
Modern Language Journal, 25(5), 405–417. doi:10.1111/j.1540-4781.1941.
tb03031.x
Kroll, J. F. & Stewart, E. (1994). Category interference in translation and
picture naming: Evidence for asymmetric connections between
bilingual memory representations. Journal of Memory and
Language, 33, 149–174.
Lemhöfer, K., Dijkstra, A. & Michel, M. (2004). Three languages, one ECHO:
Cognate effects in trilingual word recognition. Language and
Cognitive Processes, 19, 585–611.
202
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
Midgley, K. J., Holcomb, P. J. & Grainger, J. (2011). Effects of cognate status
on word comprehension in second language learners: An ERP
investigation. Journal of Cognitive Neuroscience, 23(7), 1634–1647.
doi:10.1162/jocn.2010.21463
Miller, M. M. & Farr, G. (1939). Student recognition of some SpanishEnglish cognates. The Modern Language Journal, 24(3), 216–220.
doi:10.1111/j.1540-4781.1939.tb02902.x
Morsella, E. & Miozzo, M. (2002). Evidence for a cascade model of lexical
access in speech production. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 28(3), 555–563.
Sánchez-Casas, R., Davis, C., & García-Albea, J. (1992). Bilingual lexical
processing: Exploring the cognate–noncognate distinction.
European Journal of Cognitive Psychology, 4, 293–310.
Schwieter, J. W. & Sunderman, G. (2008). Language switching in bilingual
speech production: In search of the language-specific selection
mechanism. The Mental Lexicon, 3(2), 214–238.
Schwieter, J. & Sunderman, G. (2011). Inhibitory control processes and lexical
access in trilingual speech production. Linguistic Approaches to
Bilingualism, 1(4), 391–412. doi10.1075/lab.1.4.02sch
Sherkina, M. (2003). The cognate facilitation effect in bilingual speech
processing. Toronto Working Papers in Linguistics, 21, 135–151.
Schwieter, J. (2013). Lexical inhibition in trilingual speakers. In J.-M. Tirkkonen
& E. Anttikoski (Eds.), Proceedings of the 24 th Scandinavian
Conference of Linguistics (pp. 249–260). Joensuu, Finland:
University of Eastern Finland Press. Retrieved from http://urn.fi/
URN:ISBN:978-952-61-0924-4
Sunderman, G. & Kroll, J. F. (2006). First language activation during second
language lexical processing: An investigation of lexical form,
meaning, and grammatical class. Studies in Second Language
Acquisition, 28(3), 387–422. doi:10.1017/S0272263106060177
Szubko-Sitarek, W. (2015). Multilingual lexical recognition in the mental
lexicon of third language users. Berlin: Springer. doi:10.1007/9783-642-32194-8
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
203
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
Van Hell, J. G. & De Groot, A. M. B. (2008). Sentence context modulates visual
word recognition and translation in bilinguals. Acta Psychologica,
128, 431–451.
Van Hell, J. G. & Dijkstra, T. (2002). Foreign language knowledge can
influence native language performance in exclusively native
contexts. Psychonomic Bulletin and Review, 9, 780–789.
Yudes, C., Macizo, P. & Bajo, T. (2010). Cognate effects in bilingual language
comprehension tasks. Neuroreport, 21(7), 507–512. doi:10.1097/
WNR.0b013e328338b9e1
Appendix A
Task 1: Participants will be asked to produce synonyms to the following
list of 45 words (audio recorded with the permission of each participant)
in order to elicit natural responses.
Verbs:
1.
Show up
3.
Signal
2.
4.
5.
6.
7.
8.
9.
Introduce
Research
Say
Bother
Apply for
Chat
Remind
Adjectives:
10.
Different
12.
Weak
11.
13.
14.
204
Friendly
Current
Vague
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
15.
16.
17.
18.
19.
Unlikely
Healthy
Sensitive
Funny
Weird
Nouns:
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Behaviour
Tradition
Goal
Concern
Folder
Relative
Attempt
Greeting
Review
Environment
(Fillers):
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
Walk
Shout
Teach
Wash
Travel
Goofy
Understanding
Energetic
Cautious
Introverted
Exercise
Construction
Publication
Dream
Lounge
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
205
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
Appendix B
Task 2: Participants will be asked to rate their level of acceptability of the following 60 sentences containing “grey” cognates, false cognates, preposed
adverbs/adjectives and question order in embedded clauses.
1.
Unacceptable; meaning is completely unclear; 2.
Somewhat unacceptable; meaning is somewhat unclear; 3.
Satisfactory; I probably wouldn’t say it this way, but I understand
4.
Somewhat acceptable; meaning is somewhat clear; 5.
Acceptable; meaning is completely clear; 1.
What time did your friend appear at the party last night?
2.
You have traveled a lot, so you know many different cultures.
3.
I have an investigation due on Monday for my European History class.
4.
We saw many distinct animals at the zoo on the weekend.
5.
I would like to do a Master’s degree, but I have to solicit a position
6.
The two friends sat in the coffee shop conversing for a few hours.
7.
My best friend is such a comical person; She always makes me laugh.
8.
I’m not sure how to get to your house by car, so you’ll have to indi-
9.
If you have time, could you revise my grammar before I hand in this
10.
Chocolate can make children demonstrate poor temperament.
206
I would never express myself in this way.
I am unlikely to express myself in this way.
the meaning just fine.
I might express myself in this way.
I would most likely express myself in this way.
in a program first.
cate for me.
assignment?
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
11.
What time did your friend show up at the party last night?
12.
You have traveled a lot, so you are familiar with many different cultures.
13.
I have an essay due on Monday for my European History class.
14.
We saw many different animals at the zoo on the weekend.
15.
I would like to do a Master’s degree, but I have to apply for a position in a program first.
16.
The two friends sat in the coffee shop chatting for a few hours.
17.
My best friend is such a funny person; She always makes me laugh.
18.
I’m not sure how to get to your house by car, so you’ll have to direct me.
19.
If you have time, could you check my grammar before I hand in this
assignment?
20.
Chocolate can make children demonstrate poor behaviour.
21.
The professor presented herself to the class on the first day.
22.
The student gave a very lazy description of the experiment. It was
too general.
23.
Remember me to pick up the kids after school.
24.
I didn’t like that move because the storyline didn’t have any sense.
25.
If you don’t assist all of the classes, you lose 10% in participation for
the semester.
26.
I’m preoccupied because I might not have enough money to pay for
rent next month.
27.
I didn’t say him what happened because I didn’t want him to get
upset.
28.
My new puppy is driving me crazy; I can’t support his behaviour.
29.
Part of my new job is to alphabetize all of the carpets for each of the
clients.
30.
My mom is always telling jokes; she is so gracious.
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
207
Cognate Effect and Lexical Processing in English-Spanish and Spanish-English Bilinguals
31.
The professor introduced herself to the class on the first day.
32.
The student gave a very vague description of the experiment. It was
too general.
33.
Remind me to pick up the kids after school.
34.
I didn’t like that move because the storyline didn’t make any sense.
35.
If you don’t attend all of the classes, you lose 10% in participation
for the semester.
36.
I’m worried because I might not have enough money to pay for rent
next month.
37.
I didn’t tell him what happened because I didn’t want him to get upset.
38.
My new puppy is driving me crazy; I can’t handle his behaviour.
39.
Part of my new job is to alphabetize all of the folders for each of the
clients.
40.
My mom is always telling jokes; she is so funny.
41.
We have to make more difficult this test
42.
It’s essential to write more clearly your ideas in your journal.
43.
It is more interesting the novel than the movie version.
44.
He more often eats vegetables than junk food.
45.
You have to do more regularly the exercises in order to get in shape.
46.
We have to make this test more difficult.
47.
It’s essential to write your ideas more clearly in your journal.
48.
The novel is more interesting than the movie version.
49.
He eats vegetables more often than junk food.
50.
You have to do the exercises more regularly in order to get in shape.
51.
We’re not sure where is the hotel located.
52.
Can you tell me why is there an extra charge on my credit card statement?
53.
I wonder when is she coming.
208
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
MCGREGOR
54.
He is unaware of how much is the total of his cell phone bill.
55.
Is it possible to find out who is the person serving us tonight?
56.
We’re not sure where the hotel is located.
57.
Can you tell me why there is an extra charge on my credit card statement?
58.
I wonder when she is coming.
59.
He is unaware of how much the total of his cell phone bill is.
60. Is it possible to find out who the person serving us tonight is?
LACLIL / ISSN: 2011-6721 / Vol. 9 No. 1 January-June 2016 / doi:10.5294/laclil.2016.9.1.8 / 174-209
209